Swedroe: Beware The Fed's Equities Effect

Do stocks really pop ahead of Fed meetings? Don’t bet your hard-earned money on it.

I was asked recently to comment on a study by the staff at the Federal Reserve Bank of New York that presents investors with a puzzle related to stock returns.

Researchers at the regional Fed bank found that since the Federal Open Market Committee (FOMC) began announcing its monetary policy decisions in 1994, U.S. stocks have experienced large excess returns in the 24 hours preceding these disclosures.

These abnormal returns—which occur just prior to when the policymaking body convenes its eight prescheduled meetings per year—have accounted for more than 80 percent of stock returns above the return on riskless one-month Treasury bills.

The authors found that the S&P 500 Index has, on average, increased 49 basis points in the 24 hours prior to the FOMC announcements. And while the excess return to the S&P 500 Index during the 24 hours preannouncement is, on average, 3.89 percent per year, it is just 0.89 percent on all the other days. What’s more, they found the data was statistically significant, with a t-stat of more than 4.5.

The Puzzle Intensifies

Adding to this puzzle is the fact that other major international stock indexes have also experienced similar abnormal returns before FOMC announcements. Yet this phenomenon doesn’t exist around monetary policy announcements by their own foreign central banks. Interestingly, there’s no such return pattern in either the U.S. bond markets or with the dollar in currency markets.

In their study, the authors discussed several possible explanations for the pre-FOMC announcement drift for stocks. But they could find none that were consistent with the empirical evidence. They also found that the S&P 500 doesn’t display abnormal excess returns around other major macroeconomic announcements. The study examined nine major economic releases, including new claims for unemployment, housing starts, and GDP.

As investors, what can we take away from this information? I asked Professor Kenneth French for his thoughts. Here is what he had to say:

“I think one has to be very cautious about data-mining here. The fact that 80 percent of the equity premium for 1994 to 2011 was earned between 2 p.m. on the previous day and 2 p.m. on the announcement days doesn’t mean that someone with a different microscope cannot find some other combination of events that captures 100 percent, or even 120 percent, of the premium for exactly the same years.”

“With enough time and enough degrees of freedom—and with all the events in the world as candidates there are a lot of degrees of freedom available—we should expect researchers to discover lots of patterns like this. I am not arguing that the pattern here is the result of such search, just that with people in institutions all over the place searching, someone is sure to find something once in a while,” French continued.

Random Outcome?

“And given all the search efforts and all the possible combinations, it is hard to decide whether any observed pattern is meaningful or just a random outcome. I always lean toward random, but only time will tell,” he added.

The bottom line is the pre-FOMC announcement drift is a puzzle. You should also keep in mind that, even if it occurs by chance, we should expect some of the data to have statistical significance. Remember, with a t-stat of 2—the general hurdle for statistical significance—there’s still a 5 percent chance that the outcome was a random result.

So, without a logical explanation, this remains an anomaly that you shouldn’t be trying to exploit.

Larry Swedroe is the director of research for the BAM Alliance, a community of more than 140 independent registered investment advisors throughout the country.